Analysis of spectrum characteristics for multi-attribute seismic data from shaking table test of tunnel underpass hauling sliding surface

The objective of this research is to analyze the deformation characteristics in space and the dynamic reaction of tunnel underpass hauling sliding surfaces when subjected to potential seismic activities. The response characteristics and failure mode of tunnel linings are uncovered through an analysis of their time-frequency dynamic behavior. It is thoroughly discussed that there are correlation characteristics of plectrum amplitude in multivariate data under different excitation intensities and the spectrum difference characteristics during different shaking stages. The findings indicate that the slope's structural characteristics and failure modes correspond to the shallow slip type and the deformation trend of the deep weak basement slip type. The influence of topographic bias can be disregarded in tunnel underpass hauling sliding surfaces. The axial force ( f a ) in the tunnel lining section primarily exhibits pressure, while the bending moment ( M b ) is symmetrically distributed along the lining section. Tunnels are susceptible to collapse and invert uplift damage. The dominant frequency bands of dynamic strain, dynamic soil pressure, and acceleration are mainly concentrated within the 1–15 ​Hz range. Dynamic soil pressure and acceleration have a significant correlation, whereas the dynamic strain exhibits a weak correlation with both. The dynamic strain and acceleration exhibit sensitivity in their spectrum response before and during the main shock, whereas the dynamic soil pressure shows sensitivity in its spectrum response after the main shock. Based on the spectral response differences of multi-attribute data during various shaking stages, which can present a novel approach for dynamic monitoring and early warning of seismic actions. • We obtained the time-frequency dynamic response characteristics and failure mode of tunnel lining by shaking table test. • We revealed the correlation characteristics of multivariate data spectrum amplitude and spectrum difference. • We provide a new idea for early warning of seismic according to the different sensor types evaluate the structural response.